Welcome to my website!

Bio

I am an applied mathematician developing mathematical and computational methods for representing uncertainty in Earth system models. My research lies at the intersection of dynamics, stochastic processes, and data-driven modelling, with applications primarily in ocean, atmospheric, and sea-ice systems.

I am currently a CNRS Researcher in the OPERA Team at the Institut des Géosciences de l’Environnement and Université Grenoble Alpes. Previously, I was a postdoctoral researcher with Aretha Teckentrup at the University of Edinburgh and a PhD student in the MAC-MIGS Centre for Doctoral Training. My PhD thesis, Stochastic modelling and inference of ocean transport, was supervised by Jacques Vanneste. I previously studied Applied Mathematics at the University of Edinburgh.

Research interests

My interests are at the intersection of dynamics and uncertainty. Some specific topics are:

• Stochastic parameterisation/closure
• Probabilistic forecasting and data assimilation
• Hybrid modelling and emulation

My work is applied primarily to models of the ocean 🌊 and atmosphere ༄. Most recently I am working on Arctic sea ice 🧊.

Publications

Brolly, M. T. (2026). Stochasticity and probabilistic trajectory scoring are essential for data-driven closures of chaotic systems. PNAS, 123 (35), e2609143123, doi. arXiv.

Brolly, M. T. (2025). Stochastic parameterization: The importance of nonlocality and memory. Journal of Advances in Modeling Earth Systems, 17, e2025MS005223. doi. arXiv.

Brolly, M. T. (2023). Inferring ocean transport statistics with probabilistic neural networks. Journal of Advances in Modeling Earth Systems, 15, e2023MS003718. doi. arXiv.

Brolly, M. T., Maddison, J. R., Teckentrup, A. L., & Vanneste, J. (2022). Bayesian comparison of stochastic models of dispersion. Journal of Fluid Mechanics, 944, A2. doi. arXiv.

Thesis

Brolly, M. T. (2023). Stochastic modelling and inference of ocean transport. The University of Edinburgh. doi.

News

• New paper out in PNAS!